Achieving positive net energy in a nutrient removal facility: optimizing the Ejby Molle WWTP

The Ejby Mølle wastewater treatment plant (EMWWTP) is the largest treatment facility in Odense, Denmark, and has a treatment capacity of 385.000 population equivalents (PE). VandCenter Syd (VCS) Denmark, is the provider of water and wastewater services for Odense, and operates the EMWWTP. The load to the plant is mainly of domestic origin, and it discharges to a small river with very stringent effluent requirements. The plant is currently producing on an annual average basis, effluent concentrations of 1.9 mg/L of BOD₅, 4 mg/L of total nitrogen as N, and 0.1 mg/L of total phosphorus as P.

VCS is a very progressive utility, striving to become a model for the incorporation of sustainability principles in discharging its responsibilities. To this end, it has adopted a goal of achieving CO₂ neutrality firm-wide by year 2014. Through its own efforts, VCS has been able to implement changes in the facilities (primarily through the incorporation of the CHP cogeneration scheme) and in how it operates the EMWWTP, achieving in 2011 a very impressive level of 77% energy self-sufficiency. In early 2012, VCS selected through an international tender the team of CH2M HILL and Ramboll Denmark (two global engineering firms) to execute a project aimed at identifying additional energy optimization opportunities (EOOs) at the EMWWTP that would further facilitate the achievement of their declared corporate goal.

The approach adopted by the combined utility/consultant team aimed at first identifying EOOs that relied primarily on relatively quick to implement process modifications (considering both facility re-configuration as well as modified operational strategies), which would significantly reduce electrical power consumption and/or increase power generation from cogeneration. The adopted approach also identified more complex to implement improvements that could be considered for future implementation.

The key tool in facilitating the identification, prioritization and eventual development of the selected EOOs is a whole-plant integrated model that can simulate unit process performance and operational requirements through pollutant mass balances, while also predicting the corresponding carbon and energy (electrical, thermal, and calorific) balances. To this end, the team relied on Pro2D, CH2M HILL's proprietary simulation platform that incorporates IWA's ASM and ADM modeling relationships. The simulator was calibrated successfully with historic plant operational and electrical power consumption as well as generation data from their anaerobic digestion biogas cogeneration system. With this tool, the project team was able to quickly and very thoroughly evaluate multiple unit process re-configuration scenarios as well as the impact of implementing operational modifications to existing systems, making sure that the stringent effluent requirements were always met, while measuring improvements against pre-established metrics for energy demand, energy generation, and carbon footprint.

Historical influent, effluent, and operating data was obtained from VCS. The data were obtained primarily through supervisory control and data acquisition collection along with a handful of sampling reports for primary sludge, mixed-liquor suspended solids, and dewatered solids. A supplemental sampling program was conducted in order to provide additional information required for the evaluation of the existing processing capabilities as well as for the modeling of the treatment plant. The data were consolidated and analyzed statistically to develop an understanding of facility flows, loadings, and operations. Electricity consumption data for the Ejby Mølle WWTP were also obtained, and it's summarized graphically in Figure 2. Overall, the facility was roughly 77% self-sufficient at the onset of the project.

A first Workshop was conducted at the Ejby Mølle WWTP, and was attended by management and operations staff from VCS, as well as by representatives from the consulting team. The main objectives of this initial workshop were to charter the joint VCS/Consultant team, to review the data gathering and analysis activities, to develop a ‘long list’ of EOOs, to confirm the evaluation methodology, and to establish a ‘short list’ of viable alternatives to further evaluate in greater detail. A total of 35 EOOs were identified as part of the ‘long list’, and they were grouped in three categories: treatment process modifications (18); equipment improvements (8); and energy generation improvements (9).

The following screening criteria were used to produce a ‘short list’ during the workshop:

• readily implementable;

• primarily process modifications;

• significant impact on energy profile;

• proven elsewhere.

Based on the results of preliminary mass/energy balance simulations and using the above screening criteria, the joint team selected the following EOOs for further consideration:

• CEPT: implement at the primary settling tanks (using existing ferric and polymer feed systems) to significantly enhances removal performance and reduce organic load to aeration basins. In addition, CEPT will result in higher sludge quantities to digestion and thus have a higher biogas generation potential.

• Nitrify centrate in TFs: reroute discharge of supernatant buffer tank directly to TFs for the nitrification of plant sidestreams. TF effluent would be denitrified in BNR bioreactors.

• Decommission TF and convert TF clarifiers to CEPT for wet weather treatment: treat 100% of settled primary effluent with the BNR process; blend BNR and TF clarifier effluents prior to tertiary filtration.

• Operate at shorter BNR system SRT: this should reduce bioreactor O₂ demand as well as mixing requirements while increasing biological sludge to digester, which could result in increased biogas production.

• Reduce effluent filtration operation to 12 hrs/d: bypass all or a portion of settled secondary effluent around the effluent filtration system when it has the quality to meet regulatory requirements without additional treatment.

Three additional EOOs were identified in workshop 1 as potentially of interest, recognizing however that they would require additional effort and time to fully consider and implement:

• Co-digestion with high-strength industrial waste: adding highly concentrated, readily biodegradable supplemental waste to the anaerobic digestion system as a proven method of increasing biogas production.

• Implement sidestream nitrogen removal, to reduce energy and external carbon requirements of the mainstream biological process.

• Retrofit bioreactors with diffused aeration: fine-bubble diffusion delivers oxygen to suspended growth systems more efficiently than mechanical aerators therefore decreasing the associated energy demand. This option would replace all of the mechanical aerators or provide a combination of existing aerators and fine-bubble diffusers.

Subsequent to workshop 1, supplemental sampling was conducted which enabled a better estimation of the primary clarifier efficiency. Also, it was determined that a key industrial contributor to the influent loads at the plant (the Dalum Papir paper mill) would be discontinuing its operation at the end of 2012, and thus it was removed from the influent flow and loadings analysis. The plant-wide Pro2D model was recalibrated on the basis of this updated information, which allowed for the more detailed evaluation of the short-listed EOOS identified. The following table presents a summary of the results of the evaluation of the individual viable alternatives described above. Results are presented in terms of savings in electrical power consumption, increase in cogeneration electrical output, net increase (as a percentage) as it relates to the existing total electrical power consumption, and net change in GHG emissions associated with the change in electrical power use. For this purpose, a GHG emission factor of 384 kg of carbon dioxide equivalent per megawatt hour was used, as supplied by VCS (Table 1).

Furthermore, the next table presents the results of a series of potential ‘groupings’ of these individual options into comprehensive energy optimization scenarios for the plant (Table 2).

As can be seen from these results, all of the EOOs have the potential to significantly improve the energy profile of the EMWWTP and reduce GHG emissions associated with electrical power consumption. Many of these options have already began to be readily implemented by plant staff in a ‘gradual’ approach to ascertain the full extent of their benefit and to confirm/identify potential associated risks and thus establish mitigation actions, if warranted. These results also indicate that the objective of transforming the plant into a near net-zero electrical power facility is achievable in the short term, especially by combining some of the EOOs as described and summarized in the table. Figure 3 and Figure 4 presents some of this recent data, clearly indicating that the implementation of these EOOs has resulted in the significant improvement of the energy balance at the facility, essentially demonstrating their contribution in making this facility energy self-sufficient when considering both electrical and heat production and usage.

As can be observed from Figure 3, the overall energy balance has been positive due to sale of both electricity and excess heat. The excess heat is utilized in the town wide district heating system. Some codigestion of high-strength organic industrial waste has taken place but not on a regular basis. Values for 2014 are based on a prediction after 3 month of actual data.

As can be observed from Figure 4, the plant is expected to produce a surplus of electricity in 2014. The electricity neutrality has been delayed due to installation of additional cogeneration equipment. This can be observed in the production figures from 2012 and 2013 indicating that the maximum production has been reached. A new cogeneration unit has been online since late 2013, after which electrical power generation has exceeded total electricity usage at the plant.

The following table presents similar estimates for the two facility improvement EOOs also identified as part of the project: the incorporation of a side stream treatment component to reduce the nitrogen related recycle load coming into the bioreactor as a result of the dewatering of digested sludge, and the retrofit of the existing mechanical surface aerators from the bioreactors with a fine-bubble diffused aeration system (Table 3).

These results would indicate that by combining both operational and facility improvement EOOs, it is possible to further move the EMWWTP into a net-positive energy status. Consequently, the next phase of the project consisted in the definition and incorporation of facility improvements aimed at leveraging the low-energy profile of the deammonification process for the treatment of the ammonia-rich sidestream generated from the dewatering of anaerobically digested sludge, and by the incorporation of this process as part of the mainstream nitrogen removal process at the plant. Deammonification involves nitrogen removal using partial nitritation followed by anaerobic ammonia oxidation. This is accomplished by seeding anammox bacteria from a sidestream deammonification process into the mainstream process (essentially a bioaugmentation mechanism) which is operated with alternating aerobic/anoxic conditions and in which these specialized microorganisms are retained by wasting activated sludge selectively (Figure 5). It should be noted that this mainstream process modification does not rely on deammonification pathways exclusively for all of the anticipated removal of ammonia and nitrogen. Rather, these removals occur through a combination of conventional nitrification–denitrification, nitrite shunt and deammonification. Definition of the appropriate way to incorporate these advanced processes at the EMWWTP included a comparative evaluation of process vendors, visits to facilities currently conducting sidestream and mainstream deammonification, fast-track development of design and bid documents, and construction of associated facilities. It is anticipated these systems will be operational by the spring of 2015.

Through the analysis of historical operational data, the use of an advanced mass/energy balance simulation tool, and the adoption of a collaborative workshop-based approach in identifying and evaluating alternatives, the joint VCS/Consulting Team was able to identify a series of readily implementable operational EOOs at the EMWWTP that will significantly contribute toward the Utility's adopted goal of achieving CO₂ neutrality firm-wide by year 2014. Implementation to date of some of these EOOs has already resulted in several months of a slight net-positive electrical energy balance operation, while maintaining compliance with the plant's regulated effluent requirements thus demonstrating the feasibility and real potential of the recommended optimization options. Results of this study also indicated that the implementation of deammonification processes both in the nitrogen-rich sidestream from digested sludge dewatering, and as part of the mainstream of the plant would clearly move this plant into a net-positive energy profile. As a result, a fast-track facility upgrade project has been implemented, targeting full operational capabilities by the spring of 2015.